|Session:||Session 4PP4P - Propagation Experiments and Algorithms (14b)|
|Date:||Thursday, November 09, 2006|
|Time:||15:30 - 18:30|
Transhorizon UHF Propagation over the Sea: Observations and Predictions
Siddle, D.; Gunashekar, S.D.; Warrington, E.M.
University of Leicester, UNITED KINGDOM
D.R. Siddle (email@example.com), S.D. Gunashekar (firstname.lastname@example.org) and E.M. Warrington (email@example.com), Radio Systems Research Group, Dept. of Engineering, University of Leicester, Leicester, LE1 7RH, UK
It is well known that UHF signals can travel to distances well beyond the horizon under certain atmospheric conditions. This may be beneficial to some communications systems, but for cellular networks it leads to interference which far exceeds the levels specified in the design for a small percentage of the time. As this phenomenon occurs more frequently over water, an over-the-horizon study was carried out on three paths in the British Channel Islands at 2 GHz over a period of 2 years.
Statistics are presented to show how frequently enhanced propagation occurs, for what duration and how widely it is distributed in space. Figure 1 shows a number of different percentiles of the signal strengths for the longest path, which is between Jersey and Alderney, a distance of 50 km. The median signal strength is a linear function of the effective transmitter and receiver antenna heights, which vary with the tide. The signal strength achieved for 1% of the time is 30–40 dB above the median and barely influenced by the tide. Figure 2 shows the 21 km Jersey to Sark path has a similar 1% level, demonstrating a very weak dependence on path length, compared to the behaviour of the median signal strength.
The observed signal strengths are roughly 10 dB higher than predicted by the relevant ITU Recommendation (P.1546-1), which is based on measurements at lower frequencies. Comparison with a model based solely on diffraction (ITU P.526) shows that propagation is slightly enhanced for most of the time. Consequently, it may be that two different mechanisms for enhanced propagation are at work, one frequent and one relatively infrequent.
Various strategies are being pursued in order to explain the magnitude and predict the onset of the signal enhancements. Firstly, an hourly profile of the ambient refractivity from sea level to 100 m has been estimated from meteorological records. This indicates the times when refraction is strong enough to overcome the earth’s curvature. Secondly, multivariate correlation with meteorological parameters may unearth a more complex dependence. Finally, a study of synoptic meteorological charts may show dependence on higher-level phenomena, such as frontal or anticyclonic features. Results from these analyses are presented alongside the observations and the predictions from the empirical models mentioned above.
Monitoring of Phase Fluctuations of GPS Signals in Polar Ionosphere
Shagimuratov, I.I.1; Efishov, I.I.1; Krankowski, A.2; Zakharenkova, I.E.1; Yakimova, G.A.1
1West Department of IZMIRAN, RUSSIAN FEDERATION;
2Institute of Geodesy, University of Warmia and Mazury, POLAND
Dual frequency measuremen¢ts of differential carrier phase of GPS L1 (f1=1.6 GHz) and L2 (f2=1.2 GHz) signals have been used to study the storm-time development of phase fluctuations in Antarctic region. We used GPS measurements of the International GPS service (IGS) network stations of MCM4(78¢ªS, ¬¶~80¢ª), CAS1(66¢ªS, ¬¶~80¢ª), MAV1(67¢ªS, ¬¶~71¢ª), DAV1(68¢ªN, ¬¶~75¢ª). The phase fluctuations are caused by the presence of irregularities of different scale in high latitude ionosphere. The standard GPS measurement provided global network are sampled with 30 s. The interval enables to detect the ionosphere irregularities with scale more than tenth kilometers. For analysis of data the temporal variations of differential carrier phase GPS signals along individual satellite passes was used. In this report we present an occurrence phase fluctuations with scale times 10-30 min. which were caused by large-scale irregularities with the size more than 100 km. The slow fluctuations of differential carrier phase characterize the spatial and temporal variations of TEC along individual satellite passes. The strong TEC fluctuation as an enhancement of TEC relative phone was found out in polar stations. The enhancement of TEC was bigger the 2-10 time relative background, while the TEC increase to 10-15 TECU in interval about 5-10 min. The duration of such structures was up 10-20 min. It was observed during storm as well as during moderate geomagnetic activity. We attributed the structures as occurrence of polar patches. The intensity of TEC fluctuations increased during geomagnetic activity. On the 4-5 hour satellite passes 3-6 numbers strong and some weak patches ordinary were detected. As the measure of patch activity we used the rate of TEC on 1 min interval (ROT). In report the diurnal, seasonal, latitudinal and storm-time features of occurrence of polar patches are presented.
Evaporation Duct Refractivity Profile Estimation
Levadnyi, I.; Ivanov, V.; Shalyapin, V.
Institute for Radiophysics and Electronics NASU, UKRAINE
The refractivity profile is initial data for the microwave propagation prediction models. Evaporation duct height is usually used to characterize refractivity profile. So the statistics of evaporation duct height from realistic models can help in designing marine microwave communication and radar systems. The evaporation duct height is calculated using bulk measurement of air temperature, wind speed, humidity, pressure at ship deck level and sea surface temperature. Four prevailing models: LKB(Liu-Katsaros-Businger), RSHMU (Russian State Hydro-Meteorological University) optimized by us, ECMWF (European Center for Medium range Weather Forecast) and COARE (Coupled Ocean-Atmosphere Response Experiment) were examined. The results of computation using above mentioned models were compared with the direct refractometric measurements. All measurements (meteorological and refractometric) were made by us during two expeditions. First expedition was in the Atlantic ocean from March to May in latitude 22°-32° North and longitude 52°-65° West (29 measurements). Second one was in the Indian ocean from December to February in latitude 0°-15° North and longitude 55°-80° East (94 measurements). The approximation by least square-root method was carried out to compare the direct measurements of evaporation duct height with the results of computations. The minimum square-root error is obtained for LKB model (2.59m, for negative air-sea temperature difference 2.42m), maximum - for ECMWF model(2.72m). All models overestimate low evaporation duct heights and underestimate - high values. This effect is least of all define in RSHMU model. The sensitivity of the models to the measurement accuracy of the meteorological parameters was also examined using Monte-Carlo method. The least sensitive model to measurement accuracy of the meteorological parameters is LKB model and the most sensitive is RSHMU one. Improvement of this models is limited to the finite measurement accuracy of the evaporation duct height and the meteorological parameters. As result LKB and RSHMU models can be recommended for farther use.
Wideband Terrestrial Microwave Propagation Simulations Using Interpolation in Frequency Domain
Grabner, M.; Kvicera, V.
TESTCOM, CZECH REPUBLIC
Wideband applications become still more important in terrestrial microwave communications nowadays. Propagation conditions significantly affect the quality of microwave communication systems. In order to assess the influence of inhomogeneous atmosphere on the propagation conditions, simulations of electromagnetic wave propagation are utilized. A parabolic equation method (PEM) is often applied for that purpose. This method can calculate single frequency response efficiently. If the complete transfer function of the propagation channel in a wide frequency range is required, PEM has to be applied iteratively for every single frequency in the band. Time domain methods for terrestrial propagation have been developed to solve this issue getting multi-frequency response however a computational cost in comparison with PEM is not lowered significantly by this way.
If the frequency response at one or at a few receiving locations is only needed, as it is often in the case of microwave communications, an alternative approach is to use frequency interpolation, i.e. try to find a propagation channel model. An application of the frequency interpolation method suitable for terrestrial propagation simulations will be presented in this paper. In the method, a complex received signal is modeled as the weighted sum of complex exponentials.
This model is naturally related to physics of clear-air multipath propagation, because the received signal can be considered as a vector sum of signals coming from different directions with different amplitudes and with different phases. Only several single frequency responses calculated using PEM are needed to obtain model parameters. The complex exponential model parameters are extracted by means of the matrix pencil method. An extracted model can be utilized then to calculate the channel transfer function that is impaired by an inhomogeneous distribution of refractivity between the transmitter and receiver.
Numerical examples of the application of the method will be introduced. Results will be compared with the channel transfer functions computed using repeated PEM computations. Computational cost of both methods will be compared. Robustness of the interpolation method will be discussed.
Attenuation Due to Hydrometeors at 58 GHz on 850 m Path
Kvicera, V.; Grabner, M.
TESTCOM, CZECH REPUBLIC
Frequency bands above 38 GHz will be extensively utilised by terrestrial digital fixed services because the lower bands are practically fully occupied. It means that these frequency bands will be utilised although there is a lack of experimentally obtained primary bases which are needed for realistic calculations of both quality and availability of service. Relevant ITU-R recommendation is advised to be used up to 50 GHz only. This is why TESTCOM launched the electromagnetic wave propagation research in the frequency band 58 GHz.
TESTCOM started the electromagnetic wave propagation research in the frequency band 58 GHz in December 2000. NOKIA MetroHopper equipment with transmitting frequency 57.650 GHz and V polarisation has been used. The length of the experimental path is about 850 m. The recording margin has been 30 dB due to special off-set antennas developed at TESTCOM. Since October 2002, this research has been extended to measurements at 57.250 GHz with H polarisation exactly on the same path. The original NOKIA antennas have been used and the recording margin has been 16 dB. The research is focused on attenuation due to hydrometeors. Rain attenuation data obtained from two parallel paths at 58 GHz, polarisation H and V, was processed over a 3-year period from October 2002 to September 2005.
Rain intensities have been measured at TESTCOM since February 1992 by means of a heated siphon raingauge. Rain intensity data from raingauge records was statistically processed over the same periods as was the rain attenuation data.
The cumulative distributions of attenuation due to individual hydrometeors for the first 1-year period of observation, for the second 1-year period of observation, for the third 1-year period of observation and for the entire 3-year period of observation together will be given. Influence of individual hydrometeors on attenuation at 58 GHz band will be assessed. Year-to-year variability will be shown. The cumulative distributions of rain attenuation obtained will be compared with the calculated ones in accordance with the relevant ITU-R recommendations. The results obtained will be broadly discussed.
A Neural Network Model for the Separation of Atmospheric Effects on Attenuation: Comparison of Rain Attenuation Statistics
Mallet, C.1; Barthes, L.1; Marsault, T.2
Attenuation due to the propagation of radio waves through the Earth's atmosphere plays a major role in satellite link attenuation in high frequency bands, between 10 and 50 GHz. It is caused by several types of atmospheric component: gases (oxygen and water vapour), clouds and rain. Each of these three components behaves quite differently, when considered in terms of its temporal and spatial variability. Separation of the different atmospheric contributions (also called separation effects) is an essential step for the improvement of propagation model. In fact, each of the atmospheric components is subject to temporal variations, which are individually specific; moreover, the relative contribution of each component to the total attenuation varies with frequency. Separation effects is thus necessary to dynamic modelling or for the development of frequency scaling model.
Our aim in this study is to valid a statistical model able to separate out the respective roles played by the three types of atmospheric contributor. An artificial neural network (ANN) is trained to estimate the relative contributions of the three contributors, namely gases (oxygen and water vapour) clouds and rain, to the overall attenuation measured at one or more frequencies. This training is performed in a supervised manner, and requires a wide database of inputs and corresponding targets is simulated from a wide set of atmospheric profiles, corresponding to different sets of meteorological conditions. A simulated dataset, derived from absorption and diffusion models applied to atmospheric profiles, is thus used. It is composed of total attenuation simulated at different frequencies and corresponding three relative contributions.
Different architecture are compared; results corresponding to attenuation inputs provided at following frequencies (f1=20 GHz, f2=30 GHz and f3=40 GHz) are presented. The single frequency model, with only one input (attenuation at one frequency) shows poorer performance, with systematically underestimated relative contributions in the case of heavy rain. The models with two inputs (attenuations at f1 and f2 frequencies) or three inputs (attenuations at f1, f2 and f3 frequencies) present very similar performances, and it can be seen that the addition of a third frequency brings no significant improvement to the separation. Actual measured attenuations performed during Olympus or Italsat experiment are used as inputs of the artificial neural network, and corresponding rain contribution are computed. Independent rain measurement performed during these propagation experiments allows another estimation of rain attenuation. Statistics of rain attenuation are then compared for the validation of the separation effects algorithm
On the Proper Utilization of Site Diversity Mechanism in Millimeter-Wave Band PMP Systems
Zvanovec, S.; Pechac, P.
Czech Technical University in Prague, CZECH REPUBLIC
Rain attenuation is one of the main factors constraining the range of high-reliability of point-to-multipoint systems operating in millimeter wave bands. Site diversity can be used as one of the propagation impairment mitigation techniques particularly thanks to the non-uniform spatial distribution of the rain rate (main and diversity link are not subjected to the same rain intensity at the same time). A diversity gain concept is usually used for diversity evaluations. A continuous simulation of rainfall fields, incorporating both the spatial and the temporal structure, necessary for proper understanding of point-to-multipoint system behavior, can be appropriate achieved only if supported by either adequate radar data or statistically based rainfall models. In our case the radar based rain data have been utilized as an input for system simulations. Large rain event database for Czech Republic was build for the simulations utilizing time sequences of rainfall radar images (three years data - almost 5 million of meteoradar scans).
Using space-time simulations of a point-to-multipoint system in a specific area under specific rain conditions we investigated the behavior of the system as a whole. In order to describe as much accurate as possible influences of rainfalls on the system, an outage improvement probability function has been set as the main specific factor. The outage improvement probability can be defined as the ratio between the successful trials to establish the diversity link and total number of all main link outages.
In the paper the outage improvement probability will be presented as a function of rain and PMP system parameters (see in Fig. 1). The analytical study of the PMP system behavior during rainstorms will be discussed. The proper utilization of the site diversity mechanism in the complex point-to-multipoint system scenario will be highlighted. The yield of the site diversity utilization in terms of link lengths, angular separation and rain parameters will be presented in the paper. Guidances to the utilization of site diversity in the whole PMP system will be introduced.
Comparative Analysis of Rain Intensity Estimation Algorithms
Romo, J. A.1; Fernandez Anitzine, I.1; Perez Fontan, F.2
1University of The Basque Country, SPAIN;
2University of Vigo, SPAIN
The more accurate knowledge of the prevision of the rain intensity values at any site are needed for a correct design of terrestrial and satellite radiocommunications links for services working at frequencies above 10 GHz.
This study is about the estimation of the rain intensity in the Basque Country, analyzing the results obtained when using various interpolation algorithms to rain intensity samples at 80 sites regularly distributed in its geography.
The rain intensity, not exceeded more than 0,01 % of the time of any month of the average year at every site of the study with integration period of one minute, has been used as interpolation variable. These values have been obtained by means of the cumulative long-dated (more than ten years) distributions of the rain intensities at each location.
Ten interpolation methods have been analyzed: Kriging, Inverse Distance to a Power, Minimum Curvature, Nearest Neighbour, Natural Neighbour, Polynomial Regression, Radial Basis Function, Triangulation with Linear Interpolation, Moving Average, and Local Polynomial.
Four interpolation surface groups of different sizes have been used, being the biggest the one which involves the whole Basque Country and the smallest the one which divides the Basque Country in sixteen regions. The other two interpolation surfaces divide the Basque Country in three and five parts respectively. For the election of the interpolation surfaces climatic and statistic criterions have been considered. For the determination of the tendency surfaces statistics of second order (cross correlation coefficients and joint probabilities) have been calculated to pairs of meteorological stations.
Estimations made with each of the methods and each of the interpolation surfaces have been compared. The statistical parameter Root Mean Squared Error (RMSE) has been the evaluation criterion for the various estimations made. The Cross Validation has been used to obtain de RMSE and also to get other statistics as the Mean Relative Error (MRE) and the Mean Absolute Error (MAE). Each interpolation method has each own parameters, which have been adjusted to get the least RMSE, and therefore, a better estimation of the rain intensity.
The results have been plotted on contour maps as isolines and surface maps.
Counteracting Rain Attenuation in Ka Band Communications Satellites:
A Systems Engineering Approach
Lisi, M.1; Angeletti, P.2
2European Space Agency, NETHERLANDS
The advent of Ka-band satellite communications is driven, besides the diffusion of consumer applications (such as High Definition Television broadcasting), by the increased demand for broadband networks in network centric systems. One the major drawbacks still affecting Ka-band satellite communications is that related to rain attenuation.
Rain attenuation in satellite communications systems operating at ka-band frequencies is much more severe than that usually experienced at lower frequency bands: rain attenuation at 20 GHz (Ka-band down link), for instance, is almost three times that at 11GHz (Ku-band down link).
A number of compensation techniques have been envisioned and experimented over the years. One first classification of such techniques is between ground-based and space-based solutions.
Ground-based techniques are those basically operated on ground (either in open or closed-loop configurations) and only indirectly affecting the satellite payload. Among them, Up-Link Power Control (both for transparent and regenerative payloads), End-to-End Power Control (for transparent payloads only), Adaptive Coding and/or Modulation, Site Diversity and Terrestrial Back-ups.
Space-based techniques involve an adaptive use of the payload resources to counteract rain attenuation.
Down Link Power Control is a technique suitable for both "bent pipe" and regenerative payloads. Adaptive Coding and Adaptive Modulation techniques can be used, in regenerative payload, to adapt the down-link signal characteristics to rain attenuation, independently from the up-link signal.
Adaptive down-link antennas allow a reconfiguration of the gain distribution over the coverage to compensate for higher rain attenuations in specific areas. Contoured beam antennas (in array-fed reflector configurations) can achieve such capability by incorporating variable components (such as variable power dividers) into their beam forming networks.
Active phased-array antennas (either in "array-fed reflector" or in "direct radiating array" configurations) offer an even better control of the down-link EIRP. In an active down-link antenna each radiating element (feed or subarray) is individually fed by a transmit active module (solid-state power amplifier). These antennas not only allow a fast and effective reconfiguration of the gain distribution over the coverage, but also make possible, in case of multiple-beam configurations, a full reallocation of the RF power among the radiated beams. In an active phased-array antenna all active modules (power amplifiers in these case) contribute to the formation of any single beam; this implies that the overall RF power can be dynamically shared among the beams (Power Reconfigurability). An alternative adaptive antenna configuration is the Multibeam Adaptive Antenna (MAA) concept, that allows a flexible sharing of the total power from the power amplifier modules among the beams by means of a Multiport Power Amplifier (MPA).
The previously presented adaptive compensation techniques are most often applied one at the time. The development of the technology and the highly integrated nature of today's satellite communications systems suggest combined approaches, implementing several techniques in parallel in a dynamic configuration. Such an integrated approach will achieve a very effective solution of the rain attenuation impairment, making the Ka-band spectrum fully available for broadband satellite applications and network centric systems.
Spectrum Efficiency Gains Resulting from the Implementation of Adaptive Transmit Power Control in Fixed Terrestrial Links at 38 GHz
Callaghan, S.A.1; Inglis, I.2; Hansell, P.2
1CCLRC-Rutherford Appleton Laboratory, UNITED KINGDOM;
2Aegis Systems Limited, UNITED KINGDOM
Adaptive transmit power control can be used to improve the spectrum efficiency of terrestrial point to point fixed links by limiting the transmit power to that required to maintain a constant bit error rate (BER) regardless of the propagation conditions. This results in a reduced transmit power being used during clear sky conditions, lowering the interference resulting from the ATPC link. This improves the frequency reuse factor associated with a given band and geographic area, providing a spectrum efficiency gain.
For systems operating at frequencies of above 10 GHz, the primary propagation impairment is rain. The spatio-temporal distribution of rain fields determines whether interfering links are attenuated in similar proportion to wanted links, thereby indicating if the implementation of ATPC will result in increased levels of interference. Both measured meteorological radar rain data and fractally simulated rain data were used to analyse the effectiveness of ATPC.
This paper describes a project funded by Ofcom as part of its Spectrum Efficiency Scheme. The project produced two pieces of software:
1. The planning tool, which takes an existing frequency plan for a given band and re-plans it, subject to the mix of ATPC and non-ATPC links and the type of ATPC in use. The statistics of the new plan are then calculated to estimate changes in band efficiency.
2. The analysis tool takes a plan generated by the planning tool (or by another process) and applies a sequence of rain fields, evaluating system performance as measured by outage probabilities. For each rain field, the fade on each link is calculated, which then allows the EIRP uplift to be determined for each ATPC link. Every link is then tested in turn against all interfering paths, for all rain fields, and the number of outages recorded (distinguishing between those outages directly caused by a rain fade and those outages caused by ATPC-enhanced interference).
It was found that implementation of ATPC in the 38 GHz band gives significant improvements in spectrum efficiency as measured by the increase in the number of links assigned to channel 1 (from ~50% to ~70%) and the decrease in the maximum bandwidth used (from ~300 MHz to ~180 MHz). The introduction of ATPC does give rise to a number of additional outages in the presence of intense rain (~10% increase for a frontal rain event). Adjusting the wanted to unwanted ratio (W/U) in the planning process is a more effective technique for reducing ATPC-induced outages than adjusting the fade margins or interference margin. However, it is evident that none of these band-wide mitigation techniques targets the ATPC-induced outages as effectively as a judicious choice of ATPC system parameters.
One of Possibilities to Use TEC-GPS Data for HF Links
Maltseva, O.; Barabashov, B. G.; Maltseva, O. A.; Rodionova, V. T.; Shlyupkin, A. S.
Institute of Physics Rostov State University, RUSSIAN FEDERATION
Oblique radio wave propagation in HF links is very important for many purposes, for example, for single-station location (SSL) of distant radio transmitters. Main factor that can limit the performances of SSL-method is precison of the estimation of wave arrival angles using a direction finding system. Theoretically, it is possible to provide precision ~0.01-0.1 grad but experimentally this precision is ~0.7-1 grad. That is why, an acceptable accuracy of distance determinations by modern-fingers (for example, Thomson-CST) is ~10% from ground range. Meanwhile, successful application of SSL method is possible only by means of effective adaptation of the existing empiric ionospheric models to real-time conditions along the path. Such an adaptation allows estimating limit precision connected with knowing ionospheric conditions. In this paper methods of adaptation of the IRI model by means of TEC-GPS data are used. Results of MUF and ground range D determinations are compared with results of adaptation by means of VS data using elevation angles. The root mean square error describing deviation of observational dependences of MOF (UT) and D from ones predicted by the IRI model is selected as an efficiency criterion. All estimates were fulfilled for an example of month data of MOF on Inskip - Rome path (D~1735km) provided by QinetiQ(UK) group for December 2003. In this case TEC-adaptation provided the same results as VS-adaptation: mean limit error of MUF determination seemed to be 4%. For ground range D this limit seemed to be 2%. These figures correspond to results obtained by authors earlier on the other HF links.
Important Aspects of Transhorizon Propagation at 2 Ghz over the English Channel
Gunashekar, S.D.; Siddle, D.R.; Warrington, E.M.
University of Leicester, UNITED KINGDOM
Three independent 2 GHz radio paths have been established in the British Channel Islands to study the characteristics of transhorizon UHF propagation over the sea. The paths are: Jersey to Alderney (48.5 km), Jersey to Guernsey (33.5 km) and Jersey to Sark (21 km). Each site has two antennas: the high antenna heights are at 17.5 m (Jersey), 13.0 m (Alderney), 14.0 m (Guernsey) and 13.0 m (Sark) above mean sea level, while the low antenna heights are at 14.5 m (Jersey), 10.0 m (Alderney), 10.0 m (Guernsey) and 10.0 m (Sark) above mean sea level. Signal strength measurements are made using alternately the high and the low antennas.
Through our research, we aim to explain the variation in the received signal strength both on a diurnal and seasonal scale. A detailed study of the statistical variations in the meteorological parameters that influence over-sea UHF propagation has been undertaken. Relevant propagation mechanisms that result in signal strength enhancements (e.g. low-lying evaporation ducts and higher-level ducting structures) have been identified and suitably modelled (using EM radiowave propagation assessment tools like AREPS). Statistical data related to tropospheric ducting in the region have been generated and compared to historical data (Fig.1).
Our observations reveal that signal strength enhancements occur approximately 8% of the time on the Jersey-Alderney link. During these times, there are marked departures from the mean meteorological conditions that prevail in the English Channel. For example, during signal enhancements, there is an appreciable decline in the measured wind speed accompanied by a change from unstable atmospheric conditions to stable conditions. Furthermore, the occurrence of enhanced signal strengths follows a distinct diurnal as well as seasonal pattern, occurring predominantly in the late afternoon and evening periods during spring and summer (Fig.2). With particular reference to the Jersey-Alderney path, for majority of the observations (98.7%), the high antenna signal strength is always equal to or higher than the low antenna signal strength with a mean signal strength difference of approximately +4.9 dB. Additionally, for 92% of these observations, normal (i.e. non-enhanced) signals are observed at both the high and low antennas. However, for the 8% cases of enhanced signals, there is a distinct reduction in the offset between the high and low antenna signal strengths (mean signal strength difference = +2.1 dB). In fact, overall approximately 1.3% cases of signal strength reversal (i.e. when the low antenna signal strength exceeds the high antenna signal strength) are recorded, of which almost 68% are enhanced signals (with a mean signal strength difference of about -1.5 dB), indicating a definite change in the nature of the ducting structure(s) responsible for enhanced transhorizon propagation at 2 GHz.
The various issues under investigation are of direct relevance in the planning of maritime and coastal radio communication systems operating in the UHF band, and will be input to the appropriate study group of the ITU Radiocommunication sector in order to update relevant recommendations (e.g. ITU-R P.1546).
Investigation of the Polarization Properties of Satellite Channels with Multiple Antennas
Horvath, P.; Frigyes, I.
Budapest University of Technology and Economics, HUNGARY
In contrast to a plane-wave situation electromagnetic wave polarization is a 3-dimensional phenomenon in multipath scatter environments. Recently there is much research work done in order to utilize this fact in MIMO wireless channels. It is shown that with appropriate antenna arrangement increase in capacity can be as much as 6-fold in terrestrial rich scattering environments. In this paper satellite systems are investigated from this point of view. It is shown that it is hard to achieve an increase of capacity. However, diversity gain (or the order of diversity) can be increased and in principle diversity gain of 6 or more is possible while this gain is at most 4 in the 2D (i.e. plane wave) case.
It is proved that due to the special character of a satellite link - i.e. very long free space section and short multipath section - only 2D polarization of the waves can be utilized. If only one satellite can be employed, the conventional space diversity effect cannot be utilized because the paths originating in different antennas are very highly correlated. This can be attributed to the very small apparent angle difference between the satellite antennas. Thus the capacity of a satellite MIMO channel can only be increased by a factor of 2 only, relative to the single channel case. Higher capacity could only be achieved in a multi-satellite system with appropriate inter-satellite synchronization.
On the other hand diversity can take the full advantage of the capabilities of multiple antennas if these are applied in the ground terminal. As a consequence, this type of channel is asymmetric: the downlink is a double input-, multiple output channel, the uplink is its inverse. The implications of this asymmetry in the code design is discussed.
The new results in the paper include the following contributions. We characterize the polarization scattering matrix of one random scatterer by finite-difference time-domain simulations. The results are compared to the Gaussian assumption generally used in the literature.
Furthermore, we provide by similar numerical calculations the channel characteristics for a rich, randomly scattering environment where the scatterers located close to the receiver and the transmitter uses two orthogonal polarizations. We show that, due to the scattering, polarization diversity can be used efficiently with i. e. 3 orthogonal receiving antennas.
Capacity CCDFs and BER curves for a tipical 2 transmit, 3 receive antenna space-time block code are given. Finally, simulation results for a realistic satellite-to-indoor scenario is presented. A typical office room with realistic material properties and scattering geometries is built up. A satellite using two orthogonal polarizations illuminates the room at different azimuthal and elevation angles. The receiver is placed at different points of the room with different visibility (LOS, obstructed). The simulation yields the constituent channel power delay profiles with very good resolution. In this manner, the 3x2 channel matrix is obtained directly at various points in the room. We show that in such realistic scenarios the polarization diversity is advatageous as scattered components can be observed on orthogonal polarizations. The power of these components is significant power, sometimes comparable to the power of the direct path.
Characterization of Temporary and Space Variability of the Rain Cells and Fronts and the Aplication to the Propagation of Signals
Romo, J. A.1; Fernandez Anitzine, I.1; Perez Fontan, F.2
1University of The Basque Country, SPAIN;
2University of Vigo, SPAIN
This paper presents the study of rain cells in the geographic area of the Basque Region in order to a later application in radiocomunications. It will be analized both, the space and the temporal variability of rain rates , to obtain a reliable characterization by parameters like size, speed or direction of propagation of rain cells.
The database used to develop the models and to simulate is composed by 10 years of rainfall data collected in the Basque Country. 47 meteorogical stations distributed across the Basque Country are used. These data are gathered through the rain gauge of each station. As measuring element, is used a tilted balance whose upsets are detected, and written down in the file. Therefore the data recopilated are the exact moments at which the cup of the rain gauge, whose capacity is of 1mm, upsets.
In the first place general pattern for the rain events has been identified .This pattern can be observed in the vast majority of the meteorological stations of the territory. In these stations, 75% of the characterized events can be described as stratiform type of rain. In addition, the amount of rain events of convective nature, reduces considerably as it increases the intensity of precipitation. In most of these stations the very intense rain events does not exceed the 5%.
An exhaustive study of the dynamics of the rain fronts of the territory, has been developed using the software designed. By this way, the displacement of the fronts in the horizontal direction has been observed.
For different gaps of time, the number of simultaneous events between each pair of stations is calculated, and represented.
For each pair of stations, it is analysed when a maximum in the number of simultaneous events exists, also studying the presence or not of secondary lobes in the graphs.
In addition to the analysis, the study of events with very high intensity of precipitation has been developed. The movement of these events are also represented.
Finally, this work obtains the necessary conclusions to apply the results on the design of dinamic fade countermeasures, in those radio links working at frecuencies over 10 GHz.
Subsurface Illumination and Propagation in Ground Penetrating Radar
Pennock, S.; Redfern, M.; Shan, Q.
University of Bath, UNITED KINGDOM
Ground Penetrating Radar is required to accurately resolve the positions of buried objects, both in plan position and depth. The Mapping the Underworld project (www.mappingtheunderworld.ac.uk) aims to find the positions of buried targets to an accuracy of about 3cm at depths up to 2m.
Traditional GPR places the transmitter and receiver antennas on the surface of the ground and illuminates down onto the targets whose reflections are collected in the receiver antennas. The GPR operates over a very large dynamic range, particularly when looking for deep targets buried in very lossy media.
A novel variation on the theme of GPR is to place the transmitter or receiver in the ground, typically in a buried pipe. Propagation paths are then shortened, and higher frequencies and bandwidths can be used achieving better resolution than in the traditional two path radar mode of operation. The figures show a typical cross section through a pavement and road, and the field strength predicted by ray-tracing software. In this case the transmitter is placed in the 'gas-pipe'.
The resolution of a GPR system can be improved by increasing the bandwidth of the signal used. The loss in many ground media increase dramatically with frequency. This effect can be modelled assuming the ground media to be a Drude or Lorentz dielectric with resonant frequencies that are higher than the GPR operating frequency. There are also cases where the resonant frequency can be in the typical 100MHz-1GHz operating band. The tendency for the attenuation rate (dB/m) to rise in proportion to frequency means that the higher frequencies become severely attenuated over larger distances. This further increases the dynamic range that the traditional GPR has to have in order to provide it's required resolution.
Using a step frequency or FMCW radar system with a receiver bandwidth of 5kHz and an overall noise figure of 4dB signals down to about -130dBm can be detected. If the transmitter power is 30dBm the transmitter to receiver link loss will be 160dB. Working over a 2m path suggests a maximum attenuation rate of about 80dB/m, which is typical of quite lossy soils. The losses could be as high as 320dB/m over a 0.5m path, which is typical of wet loam and clay soils at higher frequencies.
Since the transmitter is in the sub-surface the receivers will naturally monitor the signal emerging from the ground, and the system can ensure that its radiation meets the specifications for UWB radars and FCC Part 15. Measurements in highly attenuating conditions can be achieved by increasing the transmitter power and/or reducing the noise bandwidth. The interference to the GPR system and the general public can arise from another propagation routes. With the transmitter in a pipe its signal can propagate along the pipe, out of the access manholes, and then out into the open air. If the pipe is air or gas filled it will act as a leaky feeder which will gradually illuminate the lossy soil, so little interference is likely. If the pipe is water filled, it will act as waveguide as its permittivity exceeds that of the surrounding soil. In that case the interference is much more significant.
Polarization and Curvature Effect for Power Estimations in Leontovich Surface Impedance Configurations
Shtrom, A.; Soldea, D.F.; Einziger, P.D.
Technion - Israel Institute of Technology, ISRAEL
Recently, the study of near-field interactions of electromagnetic sources with highly lossy media, utilizing Leontovich surface impedance approach (T. B. A. Senior, Appl. Scien. Res., B8, 418-436, 1959; G. Tyras, Radiation and Propagation of Electromagnetic Waves, NY Academic, 1969), has led to analytic readily-interpretable expressions for the main power relations involved in such configurations, namely, source power efficiencies and Specific Absorption Rate (SAR) (D. Razansky et al., IEEE Trans. EMC, 47, 2005; D. Razansky et al., J. Appl. Phys., 95, 8298-8308, 2004; D. F. Soldea et al., J. Appl. Phys., 94, 2053-2059, 2003). These models, treating a semi-infinite absorbing medium in the vicinity of either finite (line current) or infinite (planar current sheet) sources, resulted in an explicit dependence of the power relations on both the geometrical and physical parameters and provided an effective mean for obtaining physical insight into the basic power absorption mechanisms as well as tight bounds and estimates on the power relations.
In some realistic configurations, however, including cellular phone interaction with living tissues (e.g., R. W. P. King, J. Appl. Phys., 87, 893-900, 2000; G. Lazzi et al., IEEE Trans. MTT, 52, 1853-1855, 2004), and electromagnetic hyperthermia-based treatments (A. J. Fenn et al., Int. J. Hyperthermia, 15, 45-61, 1999), the absorbing structure’s curvature and the radiation source's polarization must be incorporated into the model in order to make its implementation effective. Herein, the Leontovich surface impedance approach, applied for finite-size cylindrical absorbing medium, leads to a discrete Parseval's representation for the power relations in terms of the individual cylindrical harmonics' (spectral) power content. Asymptotic evaluation of the continuous Parseval’s identity, obtained via Watson transform (J. A. Kong, Electromagnetic Waves Theory, EMW, 2000; G. Tyras, Radiation and Propagation of Electromagnetic Waves, NY Academic, 1969), results in a closed-form expressions for the power relations involved. The expressions, obtained for both TE and TM polarizations, can be readily interpreted as for the previous (planar) case but with a curvature correction term, which depends, continuously, on the effective surface curvature and the source location. The inclusion of the curvature correction terms for both polarizations, whenever applicable, is thereby enables a better understanding and design of prototype systems, which involve electromagnetic sources closely coupled to highly lossy structures.
Bistatic Electromagnetic Scattering from a Stack of Two One-Dimensional Rough Interfaces Separating Homogeneous Media
Pinel, N1; Déchamps, N2; Bourlier, C1; De Beaucoudrey, N1; Saillard, J1; Toutain, S1
2Chemical Engin. and Applied Chemistry, CANADA
The study of electromagnetic scattering from a stack of two one-dimensional rough interfaces separating homogeneous media has a large number of applications: for example, in optics for coated surfaces, in remote sensing for the monitoring of oil spills and in the detection of buried interfaces (e.g., in sediments) using ground-penetrating radar. In this paper two methods are proposed to solve this problem :
The scattered electromagnetic field and its derivative on each interface, which are the problem unknowns, are obtained from the integral equations. To calculate them, one uses the MoM with test Dirac delta functions and pulse basis functions, which leads to a linear system of the form Z.X=b., in which X is the unknown. A direct inversion of the matrix impedance Z can be used with a limited number of unknowns on each interface, typically around 800 for a personal computer (2 GHz processor, 1 Gbit RAM). To overcome this issue, we developed an efficient and fast numerical method, called the Propagation-Inside-Layer Expansion (PILE) method, requiring less memory space and computing time comparatively to a direct inversion of Z. Its main interest that it is rigorous, with a simple formulation and with a straightforward physical interpretation. This last property relies on the fact that each block of Z is linked to a particular and quasi-independent physical process occurring during the multiple-scattering process inside the layer: local scattering on each interface and both upward and downward coupling.
Besides, we developed a very fast asymptotic method based on the Kirchhoff approximation reduced to the geometric optics approximation and valid for surfaces with large curvature radii and height standard deviations comparatively to the incident electromagnetic wavelength and for any surface statistics. The shadowing effect, important under grazing angles, is taken into account. Like the first method, from the Huygens principle, the scattered field is derived in the far field for any order of reflection of the system (the order of reflection corresponds to the number of reflections inside the slab).
Comparisons of the radar cross section obtained from our rigorous numerical method and other methods presented in the literature showed good agreement, thus validating our approach. For Gaussian statistics with Gaussian surface height correlation, the asymptotic method is compared with our benchmark method within the radar cross section. A good agreement is observed for height standard deviations larger than half the electromagnetic wavelength and for moderate surface slopes. This method is then applied to a sea covered in oil to calculate the contrast between a contaminated sea surface and a clean sea surface.
Electromagnetic Scattering for Large Problems with the Physical Optics - Multilevel - Binary Space Partitioning (PO-ML-BSP) Algorithm
Martinez Lorenzo, J. A.1; Gutierrez Meana, J.1; Las Heras A., F.1; Garcia Pino, A.2
1Universidad de Oviedo, SPAIN;
2Universidad de Vigo, SPAIN
This paper presents Physical Optics-Multilevel-Binary Space Partitioning algorithm (PO-ML-BSP) for calculating electromagnetic field levels in large scattering problems. In such problems, high frequency techniques usually replace full wave techniques due to their computational cost.
PO-ML-BSP uses the Physical Optics (PO) approximation, based on the estimation of induced currents, so the space is divided into lit and shadow regions. Consequently, it is very important to distinguish between viewed and hidden facets of the geometry for a considered point of view. The Binary Space Partitioning (BSP) can provide this goal in an efficient way for a small number of facets. This is the reason to propose the utilization of a Multi-Level scheme to manage large number of facets. The idea has its origin in Space Volumetric Partitioning (SVP), where the space is split into small geometric domains called voxels.
In the first step, BSP algorithm is applied. There are two phases to consider: first, a visibility binary tree is built and second, its nodes (facets) are sorted with respect to a line of sight. The tree does not depend on the location of the source or the view point, so it only needs to be built once. Then, the tree is examined for each view point to set possible visible facets (the final distinction between viewed and hidden facets will be made by using the painter’s algorithm).
Afterwards, the Multi-Level scheme creates the first-level voxels by dividing the space into domains which contain groups of triangles. Such voxels are sorted by building a BSP tree. Each one of the first-level voxels can be split into subvoxels called second-level voxels. A new BSP tree is associated with each voxel of the first-level to sort their corresponding second-level voxels. Appropriate number of levels must be chosen.
PO-ML-BSP is faster than PO-BSP because the illumination and reflections are computed voxel by voxel instead of facet by facet.
The PO-ML-BSP method has been used to compute monostatic Radar Cross Section (RCS) of two spheres. Configurations 1 and 2 are proposed (2 and 3 levels respectively). Figure 1 represents the obtained RSC as a function of frequency. The PO-ML-BSP results well to those provided by MoM and PO-BSP for high frequency values. Figure 2 presents the reduction of computational time for the Multi-Level scheme.
An Efficient Ray-Tracing Image Removal Algorithm for Indoor 3D Radiowave Propagation Modelling
Diskin, J.; Brennan, C.
Dublin City University, IRELAND
The proposed visibility algorithm does not require the use of pixels or voxels and can be used in indoor environments where multiple transmissions of a ray occur. It is assumed that surfaces of objects in the environment are regular rectangles. A projection of these surfaces from the viewpoint of the transmitter is created on a projection plane. This projection is composed of several overlapping 2D polygons which are themselves projections of the reflecting surfaces. Only the intersection and corner points of each projected surface have to be considered for ascertaining the layers of shadowing of the projection plane. These points are representative of the boundaries and interiors of each projected polygon. The algorithm judges whether the projection plane is entirely shadowed by the requisite number of facets by computing the number of layers covering all polygon intersection and corner points. If the plane is covered by a number of layers of shadowing equal or greater to the maximum number of transmissions allowed in the ray trace, the first order images associated with surfaces behind this projection plane are eliminated. This has the effect of reducing the number of images and speeding up the simulation. The algorithm only has to check the level of shadowing at key points on the projection. Initial results show a 23% decrease in execution time for ray tracing simulations involving 4th order reflections and transmissions in a typical indoor environment.
Analysis of Methods of Phase Unwrapping for Creation of DEM
Darizhapov, D.; Leonov, A.
Buryat Scientific Centre, the Siberian Branch of the Russian Academy of Science, RUSSIAN FEDERATION
The purposes of this research are:
- To conduct the analysis of existing methods;
- To create a method to simplify the calculation of phase unwrapping.
Remote sensing of the earth surface with the help of SAR is now one of the most popular techniques through which digital models of a relief are created. One of methods of SAR data processing is interferometry processing, which takes into account not only the amplitude of a signal, but also its phase component. One of the key stages in the interferometry processing of images is the so-called phase unwrapping, or the resolution of phase ambiguity in interferograms. To establish the 3D topographic contours of an area, it is necessary to know continuous change of a phase. The following methods of phase unwrapping have been analysed: without-weight phase unwrapping, Pikard’s method of iterations, the connected gradients method and growing pixels method.
Currently, work on the optimization of local algorithms of phase unwrapping is being carried out. The necessity for optimization is caused by automation processing, which requires the minimal user’s intervention in the work of the algorithm and, probably, a smaller number of the initial tasks.
A New Approach to Model Boundary Conditions in Finite Element Model of Deep Brain Stimulation
Walckiers, G.1; Pollo, C.2; Mosig, J.R.1
1Laboratoire d'Electromagnetisme et d'Acoustique, Ecole Polytechnique Federale de Lausanne, SWITZERLAND;
2Departement de Neurochirurgie, Centre Hospitalier Universitaire Vaudois, SWITZERLAND
Deep brain stimulation (DBS) is a well known technique used to treat movement disorders such as Parkinson Disease or epilepsy. However the mechanism(s) involved in this treatment remains unclear. The knowledge of the electric field released by the DBS electrode's contact can bring important information on the areas affected by the stimulation.
On the Use of Remote-Sensing Data in the Estimation of Ground Conductivity
Ruiz, R.1; Riera, J.M.2
1Instituto Politecnico Nacional, MEXICO;
2Universidad Politecnica de Madrid, SPAIN
Surface wave propagation is characteristic of lower frequency bands, as LF and MF, and allows the radio coverage to extremely large areas from a single high-power transmitter. The use of these bands has gained interest in the recent years with the definition of broadcasting standards, in particular DRM and IBOC, able to provide one or several high-quality audio services and complementary data services using the RF channels currently employed for analog AM broadcasting. Having in mind that the bandwidth of this channels is typically of 9-10 kHz, it is clear that these new systems are extremely efficient in the use of the spectrum with a minimum infrastructure.
The planning of these systems requires the characterisation of ground conductivity, as it determines the attenuation of the surface wave. Accurate estimation of conductivity can be performed by taking measurements of the field intensity around a transmitter of known characteristics. However, this technique can not be employed in areas with difficult access or far from broadcast stations. On the other hand, the cost of the measurement campaigns is not small. In some countries it is not possible to make measurements all over the territory, because of their large extension, the existence of areas with scarce population and insufficient funds. Coincidentally, because of these characteristics, the deployment of modern broadcasting networks operating in LF and MF bands can be an important asset for this countries. The main alternative in these cases is the use of geological and edaphological data to identify areas of similar characteristics, for which records of conductivity measurements exist.
In this paper, a research work carried out in Mexico is described. Its objective is the investigation of the use of remote sensing data and techniques to improve the estimation of ground conductivity, in areas where no measurements are available or can be made. To start with, measurements were taken in some particular areas so that, when combined with other measurements previously available, the total set of records of accurate data corresponds to several thousand km2. These data can be used as a reference to check the results of estimations performed with different data.
In the course of this work, data of many different kinds has been used: Geological, vegetation, humidity, edaphological, and LANSAT remote sensing data (6 bands). Since data have many different origins and formats, georeferencing has been an important issue.
It must be noted that remote sensing data by itself could hardly be a consistent indicator of conductivity, as conductivity is determined by the composition of the soil to a depth of several meters. The interest of this research is in the use of this kind of data as complementary to others. The advantage of remote sensing data is that it is available for all the territory with similar accuracy, independently of the degree of development or the population density. This is not always the case for other type of data.
The main results of this work will be described in detail in the paper. One of the most significant results is that signatures derived from remote sensing data can be useful in the estimation of conductivity, though the bands employed in these signatures should be carefully selected, as some bands present wide variations in areas with similar conductivity values.
A Reformulation of the Scattering Problem for an Aperture in a Perfect
Electric Conductor - a Hybrid Method Approach for Scattering Problems
Saab Bofors Dynamics AB, SWEDEN
In this paper a scattering problem is treated. The scattering object typically consists of a PEC (perfect electric conductor) with some kind of "disturbance" located at the surface. (The disturbance is "known" i.e. we are not treating an inverse problem.) This "disturbance" can e.g. consist of an integrated conformal antenna or some kind of composite structure, a cavity or some other type of irregularity. We simply hereafter refer to this irregularity as an aperture. A reference problem associated with the actual (real) problem is formulated; the reference problem consists of the same scattering object apart from the aperture, which is replaced by a PEC; the reference problem is thus a pure exterior PEC scattering problem. Using integral representations for the electromagnetic fields, equivalence principles and superposition, a third hypothetical "perturbed" problem is created. This problem can be interpreted as the perturbation caused by the aperture although it should be emphasised that no straightforward superposition can be made since the problems are different (with partly different domains and partly different boundary conditions). The unknowns (equivalent sources) in the IE (integral equation) of this resulting "perturbed" problem are the differences between the (equivalent) surface currents of the actual problem and the reference problem. The excitation (driving term) in the resulting problem is a tangential electric field (i.e. magnetic current) which originates from the PEC surface current of the reference problem; the excitation only exists in the location corresponding to the aperture of the actual problem. This suggests that the excitation in the IE of the "perturbed" problem can be calculated using asymptotic methods, possibly even by the physical optics approximation if it is judged to be accurate enough on the domain of the aperture. The "perturbed" scattering problem is an antenna type of problem and the (equivalent) currents of the "perturbed" problem decay away from the aperture just as in the case of an antenna problem, implicating that the computational domain can be truncated for a "small" aperture on a "large" structure. The scattering of the actual problem is given by the scattering of the reference problem (which for a large object preferably is calculated using an asymptotic high frequency method) plus the scattering from the "perturbed" problem. Thus different methods can be used to solve the reference problem and the perturbed problem implicating that a hybrid solution can be obtained. As a hybrid method approach, this method differs from previously presented hybrid approaches (to the best of the author knowledge). It is actually a decomposition of the original problem into two different problems and provides further insight into scattering problems.
A Study of the Validation of RF Energy Specific Absorption Rates for Simulations of Anatomically Correct Head FDTD Simulations and Truncated DASY4 Standard Equipment Measurements
Panagamuwa, C.; Edwards, R.; Whittow, W.; McEvoy, P.; Vardaxoglou, J.
Loughborough University, UNITED KINGDOM
Recent advances in mobile terminals have seen an increase in the number of Personal Data Assistant (PDA) type devices with mobile communications capabilities. With large viewing screens and `MS Windows` like operating systems, these devices are intended for use in front of the face, compared to the standard handsets used on the side of the head. Of significance is that RF energy from such devices has its source normal to the face and may irradiate the eyes (thought to be particularly sensitive), to a greater degree than has previously been the case. Past research by the authors  has shown that in this scenario certain metallic spectacles can increase the Specific Absorption Rate (SAR) in the eyes. However, SAR measurement for standards purposes is generally done using a DASY4  phantom. Whilst in  an anatomically correct head was used for simulation the DASY4 is homogenous and was intended for measuring handsets placed next to the ear and not to the face. In this paper propose a method to allow measurements form the DASY4 to be used to corroborate results form simulation and vice versa.
A section of the head will be removed from the back, placed face-down and filled with tissue simulating liquid. The DASY4 using an E-Field probe will scan the face region and calculate the SAR. The probe's ability to access the face is determined by the volume of head removed; removing a larger section allows more of the face to be accessed. In the standard configuration the probe is used normal to the inner surface of the phantom (or within 30° degrees of orthogonal) at the point of measurement. We also investigate the depth of brain tissue simulating liquids require for proper comparisons. Simulations were (ongoing) carried out using both in-house FDTD code and SPEAG’s SEMCAD-X  to determine the effects of removing different amounts of the head from the back. Figure 1 shows at 1.8GHz up to 10cm can be safely removed from the back of the head without a noticeable change to the Local SAR.
10cm can be safely removed from the back of the head without a noticeable change to the Local SAR. Following these results, a section 6cm deep has been decided up on as the optimal length for removal. As shown in Figure 2, it preserves the ears and allows for a large area of the face to be scanned by the probe. Simulations show that when this section of the head is removed, the 1g and 10g SAR values deviate by about +0.3% from the values obtained with the full head.
1. W.G. Whittow and R.M Edwards, "A study of changes to specific absorption rates in the human eye close to perfectly conducting spectacles within the radio frequency range 1.5 to 3.0GHz", IEEE Transactions on Antennas and Propagation, vol. 52, no. 12 pp. 3207-3212, 2004 2. Schmid & Partner Engineering AG, DASY4 Manual v4.1, March 2003 3. Schmid & Partner Engineering AG, SEMCAD-X Reference Manual, December 2005
FDTD/FVTD Methods and Hybrid Schemes Applied to Reverberation Chambers Studies
Lallechere, S.; Girard, S.; Bonnet, P.; Vernet, R.; Paladian, F.
LASMEA - UMR 6602 CNRS/UBP, FRANCE
This paper deals with temporal modelling of a Mode Stirred Reverberation Chamber (MSRC). Results obtained from FD and FV will be compared to those simulated by a hybrid FD/FV formulation on practical cases involving Reverberation Chamber (RC).
I. Numerical Models in MSRC Simulations
Basis principles of FDTD and FVTD will be briefly detailed. A hybrid FD/FV technique will be presented . Because of a high quality factor Q (i.e.: low losses’ levels in MSRC), electromagnetic fields generated inside cavities will vanish after an important number of iterations. Techniques have already been implemented to take into account losses  with FDTD.
II. Geometrical Considerations
The first example that will be presented will underline the importance of geometrical design of the mechanical stirrer of LASMEA RC. For FD simulations, stirrer generation may be improved by using discrete geometry . Fig.2 clearly shows the effect of non-conformity on various numerical results. Finally, using the commercial software FEKO, frequency domain transformation of previous temporal data will be commented in order to properly validate the impact of geometrical conformity.
Previous test configuration remains a favourable case for FDTD method because of the quite conformal description of stirrer in its initial position and the simplicity of objects (metallic plates). Simulations of an automotive device will be achieved. On Fig.3, relevance of VF/FEKO comparisons appears naturally due to the capacity of both methods to handle unstructured meshes. The commercial software Altair HyperMesh is used in this paper for designing all unstructured models. Results on Fig.4 will be completed by hybrid FD/FV ones in the final paper.
III. MSRC Temporal Simulations
Based upon data obtained from our FD and FV methods, results will be compared on equivalent devices. Numerical results will enable to distinguish advantages of methods from various criteria (CPU time, results quality and accuracy) by bringing a complete description of a MSRC problem.
This paper will show the contribution of time domain formulations for cavity problems. Impact of geometrical designs will be underlined by comparing various methods. Finally, we will lay emphasis on advantages of a hybrid FD/FV method on numerical devices involved in typical RC studies.
 P. Bonnet, et al., Finite-Volume time domain method, Time Domain Electromagnetics, S. M. Rao, Ed. San Diego, CA: Academic, 1999, ch. 9.
Asymptotic Calculation of the Scattered Electric Field from a Finite Rectangular Plate Using an Enhanced Stationary Phase Method (SPM) Approximation
Moschovitis, C.G.1; Papkelis, E.G.1; Anastassiu, H.T.2; Karakatselos, K.T.1; Ouranos, I.C.1; Freskas, N.G.1; Frangos, P.V.1
1National Technical University of Athens, GREECE;
2Hellenic Aerospace Industry, GREECE
An electromagnetic (EM) wave with wavevector ki is assumed to be incident on a perfectly conducting rectangular plate of finite dimensions. We are interested in calculating the vector potential at an observation point R(x,y,z), which lies in the near, intermediate (Fresnel) or the far (Fraunhofer) field region. We consider the above problem as a prerequisite in order to model e.g. propagation in an urban outdoor environment, which consists of three dimensional scatterers (walls). This scenario pertains to modern high frequency communication wireless networks, from GSM and UMTS to Wi-Fi and Wi-Max technologies. Operating at the frequency of 1GHz or higher, scatterers that appear in the above networks are considered to be electrically large,and current density may be calculated with good accuracy using the physical optics (P.O) approximation. Then, the vector potential at the observation point R(x,y,z) is given by the following equation:
Modifying appropriately the above equation we can apply SPM approximations resulting in calculating the vector potential and eventually the total electric field at the observation point R(x,y,z). The novel, important feature of our approach is the inclusion of the edges contribution to the resulting asymptotic expressions, which has not been documented in the literature for a double integral.
Furthermore, in order to check the accuracy of our asymptotic calculations, standard (e.g. Gaussian) numerical integration was used to compare the results. Due to the complexity of the functions on which SPM is applied, the calculations were carried out using MATLAB’s symbolic toolbox. A simulating application was also developed for the asymptotic SPM calculations. Results derived from MATLAB calculations constitute an aggregation of complicated formulas. Assuming an appropriate set of simulation parameters, the total electric field was calculated for distances in the Far Field, Fresnel Region and Near Field area. We compared SPM method and numerical integration for the frequency of 1GHz and for rectangular plates with side length equal to 20ë, 40ë, 60ë and 80ë. Both the elevation and azimuth angles of incidence were assumed equal to 45 degrees. Numerical results are shown for a rectangular plate of dimension 20ë in the Near Field Area (r=25m) and the Fresnel area (r=100m). These numerical results show satisfactory agreement between the two methods of computation.
Even though rather complicated mathematical formulas are
involved in the proposed SPM method, it is very important that this
technique is much faster than the numerical integration (namely,
SPM in this case is about 40 times faster). We can very easily
understand how important this is e.g. for a propagation problem in an
urban outdoor environment, in which case many scatterers (walls)
and multiple reflection phenomena are present. Furthermore, by
increasing the dimensions of the scatterer, we are effectively
increasing the frequency, thus reducing the error. The SPM method
presented here appears to be very attractive for the calculation of
vector potential and electric field in various radio propagation simulation tools.
Analytic Tropospheric Ray-Tracing Model for Constant Refractivity Gradient Profiles
VALTR, P.; PECHAC, P.
Czech Technical University in Prague, CZECH REPUBLIC
Prediction methods in tropospheric electromagnetic wave propagation are mostly treated by methods based on ray-tracing and parabolic equation. While parabolic equation approach represents a full-wave method of wave equation solution, ray-tracing methods take use of ray-concept of physical optics. Parabolic equation represents powerful numerical technique taking account of refractive properties of atmosphere and reflection/diffraction terrain effects. Although ray-tracing is not capable to give exact field strength/phase in general case, it provides useful tool to illustrate the atmospheric refraction effects. The main advantage of the ray-tracing concept is the relative simplicity of implementation. The aim of this paper is to introduce ray-based method of precise field amplitude/phase computation considering refractive properties of atmosphere represented by refractive index gradient.
Based on analytic solution of ray equation for constant refractivity gradient, this method is capable of tracing direct and ground-reflected ray to receive amplitude/phase at the point of the receiver. The usability of this method is limited by the requirement of line-of-sight existence between transmitter and receiver and by simple terrain scenario. Fig. 1 (left) shows direct and ground-reflected ray between transmitter and receivers at several heights under standard refractivity conditions. Thanks to the constant refractivity gradient in the whole height range these two rays are sufficient to get the value of amplitude/phase at the receiver. Fig. 1 (right) shows comparison of propagation loss obtained by the ray-tracing model and parabolic equation method.
A Method for Predicting the Scattering by a Lossy Wedge in Automotive Radio Propagation Environments
Riccio, G.; Di Napoli, D.; Ferrara, F.; Gennarelli, C.
University of Salerno, ITALY
In recent years, ray tracing has emerged as the preferred method for the coverage prediction of wireless networks and, in this context, much effort is doing to find practical high-frequency solutions for the scattering by lossy dielectric wedges. According to GTD, the total field at the observation point in the space surrounding the wedge is given in terms of direct, reflected, transmitted and diffracted contributions. Unfortunately, the derivation of an exact analytical solution for the diffracted field is usually difficult, limited to some specific wedge angles and material types. Moreover, the numerical efficiency is strongly reduced when the expression requires the calculus of special integral functions. Therefore, the derivation of approximate "practical" solutions, which are efficient, easy to handle and to implement in a computer code, becomes very appealing from the engineering point of view. Uniform Asymptotic Physical Optics (UAPO) solutions have been recently developed for solving many diffraction problems efficiently (see  as reference for the approach). They have been obtained by using a PO approximation of the equivalent surface current densities induced by an incident field on the illuminated faces of the structure and by accomplishing a uniform asymptotic evaluation of the corresponding radiation integral. The resulting analytical expressions are given in terms of the UTD transition function . Every time such solutions have been applied, they are resulted to be simple to implement in an efficient computer code and accurate when compared with numerical and measured data available in literature. The goal of this work is to propose a method for predicting the scattering by a penetrable wedge, which is here treated as a junction of two nonmagnetic lossy dielectric slabs having distinct geometric and electric characteristics, when a linearly polarized plane wave is orthogonally incident on it. The approach is based on the extension of the Burnside and Burgener's method  to determine the GO response of the two lossy material layers forming the wedge, and on the construction of the corresponding UAPO solution for predicting the diffraction contribution. In particular, since each slab is a lossy material, it must be taken into account that the transmitted field in the slab attenuates in a direction perpendicular to the incident face, whereas the phase propagates in a direction depending on the incident angle and the constitutive parameters of the slab. The presence of the other face of the slab originates multiple reflection and transmission contributions concurring to the evaluation of the reflection and transmission coefficients.  C.Gennarelli, G.Pelosi, G.Riccio, G.Toso, "Diffraction by an anisotropic dielectric half-plane: a uniform asymptotic PO solution", IEEE Trans. Antennas Propagat., vol. 49, no. 12, pp. 1624-1627, 2001.  R.G.Kouyoumjian, P.H.Pathak, "A uniform geometrical theory of diffraction for an edge in a perfect conducting surface," Proc. IEEE, vol. 62, no. 11, pp. 1448-1461, 1974.  W.D.Burnside, K.W.Burgener, "High frequency scattering by a thin lossless dielectric slab," IEEE Trans. Antennas Propagat., vol. AP-31, no. 1, pp. 104-110, 1983.
A First Approach to Equivalent Sources of 2D Complex Beams Solutions
Mahillo-Isla, R.; Gonzalez-Morales, M.J.; Dehesa-Martinez, C.
Universidad de Valladolid, SPAIN
One of the problems of the complex source method for beams generation is the fact that the sources of these fields are expressed as Dirac delta functions of complex displacements in the cartesian coordinates. This has caused several discussions about how a complex beam can be obtained since it is not clear what Dirac delta functions of complex displacement represent in the real space. There are two ways of tackling this problem.
The first one is to evaluate the corresponding field jump so as to obtain the currents that generate the field. The other one is to obtain the angular plane wave spectrum of the field and derive the currents from it, since currents and plane wave spectrum are closely related. This work presents a first approach to both methods.
We start from the field generated by the 2D complex point source method which is called complex beam [Fel76, GGD97]. The time dependence is assumed to be e−iwt. We analyze the case of a TE polarized beam radiated by electric currents (the problem with TM polarization is a dual problem by considering magnetic currents). Without any loss of generality [GGD97], only one of the coordinates of the source is complex displaced (y). Then, the electric field generated, in terms of the complex distance R = (x2 + (y − ib)2)1/2 is:
A suitable election of the branch cut of R(x, b)|y=0 for the branch points x = ±b is a branch cut the line segment joining x = ±b so that E is continuous in the x axis except the cut.
The first method consist in evaluating the electric density current through the tangential magnetic field jump between both sides of the cut. The second method consist in obtianing the density current from the plane wave spectrum, as they are a Fourier Transform pair [Cle96]. The second way is not so straightforward, but it shows some spectral interpretations. In this case, the plane wave spectrum function has also branch points and the evaluation of the Fourier transform is not easy, and suitable branch cuts will have to be chosen.
When applying High Frequency Far Field and paraxial approximations to (1), a Gaussian beam arises [Fel76, GGD97]. Gaussian beam solutions can be treated in the same way as the more general complex beam solutions. So, comparisons between the currents calculated in both cases can be done in order to track approximations. This work will be also presented.
[Cle96] P. C. Clemmow. The Plane Wave Spectrum Representation of Electromagnetic Waves. Oxford University Press & IEEE Press, Oxford & New Jersey, 1 reissued edition, 1996.
[Fel76] L. B. Felsen. Complex source point solutions of the field equations and their relations to the propagation and scattering of gaussian beams. Symposia Mathematica, 18:39-56, 1976.
[GGD97] E. Gago-Ribas, M. J. Gonzalez-Morales and C. Dehesa-Martinez. Analytical parametrization of a 2D real propagation space in terms of complex electromagnetic beams. IEICE Transactions on electronics, E80-C(11):1434-1439,
Study of Planar Transmission Lines Printed on a Ferroelectric Thin Film: Optimum Tunability and Figure of Merit
Moussavou, A.-G.1; Sauleau, R.1; Mahdjoubi, K.1; Deputier, S.2; Guilloux-Viry, M.2; Perrin, A.2
1IETR (Institut d'Electronique et des Télécommunications de Rennes)-UMR CNRS 6164/Université de Renne, FRANCE;
2Unité Sciences Chimiques de Rennes-UMR CNRS 6226/Université de Rennes 1, FRANCE
Evolution of the needs in general public telecommunications has led to intensive researches on microwave tunable devices. Among the investigated solutions to obtain agility (varactors, micro-electro-mechanical systems, agile materials), the use of ferroelectric thin films are promising for high frequency applications and planar circuits. Ferroelectrics materials are characterized by an electric field dependent dielectric permittivity and thus provide the possibility of tuning.
The aim of our work is to make a comparative study of the tunability of standard transmission lines, i.e., microstrip line (MS), coplanar waveguide (CPW) and coplanar strips (CS), printed on a ferroelectric thin film. The substrates considered in our paper are sapphire (εr = 10) and lanthanum aluminate LAO (εr = 24). The tunable range of the ferroelectric film permittivity εrf is taken from 700 to 500. The ferroelectric material can belong to the families of BST (BaxSr1-xTiO3) or KTN (KTaxNb1-xO3). The utilised electromagnetic simulator is the Ansoft’s 3D solver, HFSS.
Firstly we are interested in the tunability of different transmission lines. We apply the usual definition of the tunability (T) in terms of effective permittivity (εeff) of the lines: Tunability (%) = 100*(εeffmax - εeffmax/εeffmax. Simulations of transmission lines show that for a given value of the characteristic impedance (Zc = 50 Ω on sapphire), the MS line has much less tunability (T = 2 %) than the uni-planar configurations (T ~ 16 % for a gap width g = 0.030 mm). For the CPW, the tunability increases by 15 % when Zc varies from 40 to 60 Ω. Moreover, T is also a function of the substrate permittivity. For example, the tunability of uni-planar transmission lines decreases by 25 % when the sapphire substrate is replaced by LAO substrate.
Secondly, we investigate the insertion loss (IL) and figure of merit M (M = T/IL) of these different configurations. The MS configuration revels to be the less loss transmission line among the three, but has the lowest figure of merit (5 %/dB on sapphire substrate). Strengly enough, the uni-planar configurations have the more losses, but a much better figure of merit (6.3 and 12.2 %/dB for CPW and CS, respectively) due to their better tunability. Losses of the uni-planar lines become crippling for a loss tangent of the ferroelectric film > 0.01 but by increasing the gap g, the figure of merit increases up to a maximum value (8.7 and 15.5 %/dB for CPW and CS, respectively).
The objective of this paper is to demonstrate that, among the standard printed transmission lines, the uni-planar ones are the most tunable thanks to their electric field configuration which allows a larger impact on the ferroelectric material and on their effective parameters (Zc, εeff). But the counterpart of this, is a higher loss. Hence, the advantage of the figure of merit M is that, it represents a good trade-off between tunability and loss. The utilisation of M factor allows to conclude that the best configuration is the uni-planar transmission line which has the least metallic loss.
Inverted PI Dielectric Waveguide for Microwave Devices Design in Metallic Rectangular Waveguides Partially Filled with Dielectrics: Theoretical and Experimental Confirmation
Rodríguez García, J.; Fernández Fernández, S.; García Granda, M.; Pozo Ayuso, D. F.
University of Oviedo, SPAIN
In the present paper, the authors present a method for analyzing the complete electromagnetic scattering caused by cascade abrupt discontinuities in shielded dielectric waveguides with arbitrary sectional geometries and index profile functions. Dielectric configurations in rectangular and circular dielectric waveguides at microwave frequencies, as well as in planar and channel integrated optical waveguides are considered. The theoretical model imposes perfectly conducting surfaces as boundary conditions so that the dielectrics are enclosed by perfectly conducting walls. The cross section of the resulting guiding system is covered by a grid of pixels, taking the refraction index inside each pixel as uniform. This method is focused on a new concept of the scattering matrix; consequently, it is called Generalized Scattering Matrix Method (GSMM). In order to obtain the Generalized Scattering Matrix (GSM) of any cascaded set of abrupt dielectric discontinuities at any frequency, the Generalized Telegraphist Equations (GTE) formulation  and the Modal Matching Technique (MMT) [2-3] were extended for planar and channel integrated optical waveguides. The accuracy and effectiveness of this method was analyzed at microwave and optical frequencies. We demonstrate the possibility of designing microwave devices introducing dielectrics structures in conventional metallic rectangular waveguides. Power dividers, filters, tuners and electromagnetic band gaps devices can be built in a cheap and easy way. Theoretical and experimental results were obtained for coupled cylindrical dielectric posts, dielectric transitions, as well as for cylindrical dielectric waveguides connected abruptly and gradually. In this work, we present the new dielectric waveguide, inverted P dielectric waveguide, as an alternative for designing tunable microwave devices as well as electromagnetic band gaps configurations. Numerical and experimental results for complex scattering coefficients are given. In all cases the agreement was excellent and the possibilities of the method for designing microwave devices in rectangular metallic waveguides partially filled with dielectrics are demonstrated. The theoretical procedure was applied at optical frequencies for analyzing abrupt discontinuities in integrated planar and channel optical waveguides. We demonstrate the possibilities of the method for analysing the modal propagation characteristics of arbitrary optical waveguides as well as waveguide photonic crystals and photonic devices in optical waveguide periodical structures.
Rodríguez, J., Crespo, R. D., Fernández, S., Palacios, S. L., Virgós, J. M., "Comments on: verification of generalized telegraphists equations applied to dielectric waveguide problems", Applied Optics, 33 (1994), 356-357.
Wexler, A., "Solution of waveguide discontinuities by modal analysis", IEEE Trans. Microwave Theory Tech., 15 (1967), 508-517.
James, G.L., "On the problem of applying mode-matching techniques in analyzing conical waveguide discontinuities", IEEE Trans. Microwave Theory Tech., 31 (1983), 718-723.
Prediction of Shielding Effectiveness in Honeycomb Structure Using the Modified Design Equation
Cheong, Y.1; Lee, K.2; Hong, I.3; Yook, J.2
1Defense Agency for Technology and Quality Assurance, REPUBLIC OF KOREA;
2Dept. of Electrical and Electronics Engineering, Yonsei University, REPUBLIC OF KOREA;
3Dept. of Information and Communication Engineering, Kongju University, REPUBLIC OF KOREA
In this paper, we propose a modified shielding effectiveness (SE) design equation to obtain accurate design parameters of a honeycomb structure. With the conventional SE equation for honeycomb structures, an error occurs at frequency band, from 100MHz to 7GHz; therefore, new compensation terms were derived and introduced by analyzing the calculation results. An analysis of EMI Honeycomb structure is divided into two parts. First, a hexagonal basic lattice which honeycomb structure composed is analyzed. Analytic SE equation of waveguides (circular, rectangular, hexagonal) has been summarized and validity of each analytic equation has been demonstrated via comparison to 3-dimentional FEM simulation. The next is an analysis for infinite array of a basic lattice. Infinite array of parallel-plate waveguides is analyzed by Wiener-Hopf method. The resulting equation is shown below. The first term of the Equation (1) is the SE of unit cell of hexagonal waveguide, while the second term is the SE of infinite array of parallel-plate waveguides. (1) where, k is wavenumber, g is a transverse dimension of waveguide, d is a length of waveguide and is a angle of a incident wave. (a) (b) (g = 1.59 mm, d = 6.35 mm, α = 0¡Æ; honeycomb dimension: 20 mm by 20 mm) Fig.1 shows the comparison between numerical data and attenuation derived from the equation for hexagonal waveguide. In Fig.1 (b), it is clear that the graph of Equation (1) disagrees with FEM simulation in the low frequency region.(herein, the range of low frequency band is from 100 MHz to 10 GHz) The modified SE was obtained by adding a low frequency correction term to the conventional equation. To derive the correction term, the curve-fitting method was used. The modified SE is the following: (2) The R (R=3.18/g) is the rate of a transverse dimension of waveguide and determines when we need to supplement the third term as well as the value of the third term. (a) (b) Figure 2 SE of g change and equation (2) (d = 6.35 mm, α= 0¡Æ) Fig.2 (a) represents the values of g. Fig. 2 (b) reveals the comparison between the new SE equation and full-wave simulation results. SE of both is dependant of g variation and both converge to SE value of each g in the low frequency. The results between the designed model in the FEM domain and the modified SE equation in this paper showed good agreement and were validated through comparison with measurement data. The modified SE equation proposed in this paper can simply and accurately predict the SE of honeycomb structures generally used for EMI applications.
Propagation Modeling in Subway Tunnel Using FDTD
Nagy, L.; Nagy, L.
Budapest University of Technology, HUNGARY
The next generation mobile access network system design needs more precise characterisation of the radio channel and and needs sophisticated propagation models because of the decreasing cell sizes and of higher data rates. Especially planning coverage in tunnels and indoor spaces causes design problems without these models.
The FDTD based model is able to characterise the above scenaries and using this a costly measurement effort can be reduced using simulation methods.
In this article nearly circular subway tunnel is modeled in absence and presence of vehicle. The model can be applied for determine the field strength distribution on a cut plane of the tunnel with leaky cable excitation.
Indoor propagation problems for wideband radio systems are investigated widely and one of the today applied approach of modeling is the ray-tracing, ray-launching method. This ray methods are efficient for parallel-perpendicular scenarios but there is a common problem tracing the rays for curved surfaces. The other disadvantage of the ray methods is the difficulty in describeing the diffraction for a complex scenario. In related papers  ray tracing method introduced where bundles of rays are used to represent each "physical" wave. Monte Carlo techniques were used for the ray launching. Each bundle of rays was traced to a receiver position where reception spheres determined which rays are intercepted by the receiver.
The specific case of the straight circular tunnel can be modeled analitically as a waveguide with circular cross-section.
Each of the two previous model has disadvantage by modeling our problem, the ray tracing needs huge running time because on the curved surface reflection the number of rays in bundles has to be increased. Whilst the analitical method is not able to handle the komplex problem with the vehicle.
In our investigation the Finite Difference Time Domain method was used to analyse the 2 dimensional problem. The program in C language developed in our department and the modeling parameters were: concrete tube inner radius 3 m, vehicle size is approximately equal with our underground carriage. The simulation were performed with a point source representing the leaky cable excitation and sinusoidal 900 MHz waveform was used. The Yee cell size was 1 cm at the simulations.
Our results are showing the electric field strength distribution in the tunnel and inside of the carriage. We will compare our simulation results with measurements and also field strength results will be shown along the tunnel.
D. Didascalou, J. Maurer, and W. Wiesbeck, "Subway tunnel guided electromagnetic wave propagation at mobile communications frequencies," IEEE Trans. Antennas Propagat., vol. 49, pp. 1590–1595, 2001.
Donald G. Dudley, "Wireless Propagation in Circular Tunnels", IEEE Trans. Antennas Propagat., vol. 53, , pp. 435-441, 2005.
D. Didascalou, J. Maurer, and W. Wiesbeck, "Subway tunnel guided electromagnetic wave propagation at mobile communications frequencies," IEEE Trans. Antennas Propagat., vol. 49, pp. 1590–1595, 2001. Donald G. Dudley, "Wireless Propagation in Circular Tunnels", IEEE Trans. Antennas Propagat., vol. 53, , pp. 435-441, 2005.
Delay Time Detection for Ranging Applications Using a Tap Delay Line Equalizer
Cheeseman, B.; Huang, Y.
The University of Liverpool, UNITED KINGDOM
Wideband pulses used in ranging applications through to breast cancer detection have a common problem in that the pulse itself is not necessarily the same shape when it is received and have widely varying strengths if multiple targets are present. Travelling through a medium or when the signal reflects from wanted or unwanted objects means that the received signal is difficult to match up to what was transmitted. Common techniques to detect when the pulse is received include integration over the pulse length to detect the pulse as a region of higher than normal activity or correlation methods. The resultant signals though tend to give a fairly wide estimate of when the pulse was received and hence limited resolution of closely received pulses.
There is also a significant amount of signal masking where a stronger received pulse makes a weaker pulse difficult to detect. This paper looks at using a tap delay line equalizer to provide an estimate of a signals delay when presented with a received signal and a copy of the transmitted signal. The desired signal is designated as the received signal and the transmitted signal as the input. The LMS algorithm is used to alter the equalizer coefficients so that for the transmitted signal input the equalizer output is an approximation of the received signal.
The coefficients then provide information about the channel properties and indicate significant delay times through larger coefficient values. It is these delay times that show when a signal or multi-path copy of a signal was received. The use of a pseudo noise (PN) sequence ensures that there is a unique sequence for the equalizer to lock on to and provide accurate delay times. A simple simulation to assess the methods suitability for breast cancer detection shows the possible improvement in accurate reflected signal delay time detection over an integration method using a Gaussian pulse (Fig.1). The signal is supposed to take ~0.6ns. The integration method gives a very wide indication and is slightly skewed to the right. The equalizer gives a slightly smaller value than 0.6ns but that is accounted for by the size of the reflecting target.
A New Reverberation Chamber
Abumustafa, N.1; Huang, Y.1; Wang, Q.2; Nair, B.1
1The University of Liverpool, UNITED KINGDOM;
2Mechanical Engineering College, Shijiazhuang, CHINA
An electromagnetic Reverberation Chamber (also called Mode-Stirred Chamber, MSC) is becoming increasingly an essential tool for Electromagnetic Compatibility (EMC) testing and some other electromagnetic investigations (such as antenna diversity measurements). It is basically a conducting screened room with one or more stirrers. The stirrer is constructed from large metallic reflectors that can be ''stirred'' to different orientations in order to achieve a well stirred field inside the chamber. This reduces inhomogeneity (standing waves) specially at locations within the working/testing volume and generates a time-averaged uniform field. An ideal reverberation chamber should have the lowest possible useable frequency and a very large working volume. The Lowest Usable Frequency (LUF) of a reverberation chamber depends on many factors, such as the size of the chamber, the number and design of the stirrer and other variables which makes it harder to find an optimum design for reverberation chambers. This work aims to address the optimum design issue by examining the design of the stirrers.
A novel reverberation chamber has been designed and constructed here at the University of Liverpool. Unlike the conventional design, the new stirrers are specially designed flat panels with various 'V' cuts as shown in the figure below. The aim is to maximize the current path on the stirrer and reduce the LUF. The initial performance-based field uniformity calibration was carried out in accordance with Annex B of the IEC standard EN 61000−4−21:2003. For comparison purposes, the same chamber with conventional design (without the 'V' cuts) is also calibrated. The results show that the new design has significantly improved the averaged field uniformity and reduced the LUF by more than 20%. In addition to measurements, theoretical and numerical analysis of the chamber will be presented and should be of interest to the EMC, antennas and measurement communities.
Wavelet Analysis of the Electromagnetic Scattering from
Bandlimited Fractal Surfaces
Poirier, J.R.1; Aubert, H.2; Jaggard, D.L.3
3Complex Media laboratory,Shool of Electrical Enginnering, Philadelphia, UNITED STATES
The remote analysis of scaling properties of multi-scale (fractal) objects may be achieved from interrogation by an incident electromagnetic wave. The authors have recently reported the remote estimation of fractal descriptors of discrete self-similar laminated structures from the wavelet analysis of their impulse response [1-4]. We propose here an extension of the analysis to planar bandlimited fractal surfaces.
Bandlimited Weierstrass function consists in a suitable self-similar function which allows differentiation and provides a realistic model for describing many natural scenes . Consider a plane wave Gaussian pulse normally incident upon a metallic bandlimited Weierstrass surface. Due to the high irregularity of the surface, the near field impulse response exhibits a succession of abrupt changes or singularities in time. The Wavelet analysis of such noise-like variations is advantageously used here for detecting and characterizing the self-similarity property of the surface. The Figure shows the maxima of the wavelet transform modulus of the impulse response of a metallic and bandlimited Weierstrass surface: at a given scale, each point shown in this time-scale representation indicates the location of a maximum for the wavelet transform modulus. As indicated by the arrows, a hierarchical structure of some wavelet maxima is clearly apparent in this skeleton. This hierarchy is the "footprint" the fractal property of the surface leave in the reflection data and allows the remote estimation of the fractal descriptors of the model.
 Y. LAKSARI, H. AUBERT, D.L.JAGGARD, J.Y. TOURNERET, "Lacunarity of Fractal Superlattices: a Remote Estimation using Wavelets," IEEE Trans. Antennas Propagat., vol. AP-53, No. 4, pp. 1358 - 1363, April 2005.
 H.AUBERT, D.L.JAGGARD, "Wavelet Analysis of transients in Fractal Superlattices," IEEE Trans. Antennas Propagat., vol. AP-50, No. 3 , pp. 338-345, March 2002.
 Y. LAKSARI, H. AUBERT, D.L.JAGGARD, "The Wavelet-based Partition Function for the Remote Analysis of Discrete Self-similar Objects," Electronics Letters, vol.38, No. 14, pp.741-742, July 2002.
 A.-S. SALEH, H. AUBERT, D.L. JAGGARD, "Lacunarity of Multi-gap Fractal Superlattices Using Wavelet Analysis," Optics Communications, vol. 197/4-6, pp. 255-260, October 2001.
 D.L. Jaggard, "Fractal Electrodynamics : from super antennas to superlattices," in Fractal in Engineering, J. Lévy Véhel, E. Lutton, C. Tricot eds. Berlin, Germany, 1997, pp. 204-221.
Computations for Various Edge Configurations with the Hybrid Finite Element - Boundary Integral - Multilevel Fast Multipole - Uniform Geometrical Theory of Diffraction Method Including Double Diffraction
Tzoulis, A.1; Eibert, T.F.2
1FGAN-FHR, 53343 Wachtberg, GERMANY;
2Institute of Radio Frequency Technology, University of Stuttgart, 70550 Stuttgart, GERMANY
Recently, a powerful hybrid numerical method was introduced , combining the Finite Element Boundary Integral (FEBI) method and the Multilevel Fast Multipole Method (MLFMM) with the Uniform Geometrical Theory of Diffraction (UTD), giving full electromagnetic coupling between composite metallic/dielectric objects with arbitrary shapes and electrically large conducting objects within a common environment. In that work, single and multiple reflections on flat metallic objects, combined with single diffractions on straight metallic edges, were considered. In this contribution, the hybrid FEBI-MLFMM-UTD method is extended to double diffraction high-frequency fields on pairs of straight metallic edges, formulated with the hard and soft scalar diffraction coefficients of the UTD. The diffraction points on each pair of edges are numerically determined by a three-dimensional parametric realization of the generalized Fermat's principle. According to this, the position vectors of the diffraction points and the optical length from the source point to the observation point are expressed in terms of two parameters defined on the edges. These parameters are determined by minimizing the optical length, i.e. by setting its first derivatives with respect to the parameters to zero. The divergence factor of the double diffracted field is computed by multiplying the appropriate divergence factors of the single diffracted UTD fields on each edge for the particular case. Numerical results for various edge configurations will be presented, as for example shown in the figures below for a dielectric rod antenna with metallic mounting, radiating in front of a slot formed by two flat metallic plates. Near-field computations are accelerated in the postprocessing stage with the MLFMM using combined near-field and far-field translations , where high-frequency contributions due to the presence of electrically large objects are taken into account within each translation domain according to the hybridization of MLFMM with UTD introduced in .
|Abstracts assigned without a sequence or a sequence number beyond maximum presentation slots available:|
|39 - 347601 - A Physical Optics Solution for Bistatic RCS of Triangularly Shaped Trihedral Corners for Any Incidence and Observation Angles|